Cirrhosis is the fifth leading cause of death in the United States in individuals under the age of 65, the productive years of life. It affects men and women equally, and impacts on all races and socio- economical levels. Portal hypertension is the main complication of cirrhosis, regardless of etiology. Gastroesophageal varices and variceal hemorrhage are a direct consequence of portal hypertension and account in large part for the high mortality of cirrhosis. Non-selective beta- adrenergic blockers decrease portal pressure and have been shown to prevent the first variceal hemorrhage in patients with cirrhosis and varices. Early portal hypotensive therapy, before the patients develop varices, would be beneficial not only because it may prevent or delay the formation of varices (and variceal hemorrhage) but because it may prevent or delay the development of other complications of portal hypertension, such a ascites. This ongoing multi-center, prospective, randomized, placebo-controlled, double-blind trial was designed with the primary aim of investigating if early therapy with timolol, a non- selective beta-adrenergic blocker, can prevent or delay the development of varices in patients with cirrhosis and portal hypertension. Secondary aims will examine whether timolol prevents or delays other complications of portal hypertension such as ascites and porto-systemic encephalopathy, as well as liver transplantation or death. Patients with cirrhosis, without varices on endoscopy and with portal hypertension (portal pressure greater than 6 mmHg) are included in the study. This grant application was funded in April of 1993 and patient randomization began in August of 1993. Patient accrual took longer than originally estimated, however it is now certain that the number of 190 patients required for the study will have been randomized by the end of the current funding period (March 1998), since at the writing of this proposal 158 patients had already been randomized. In calculating sample size, we assumed a rate of development of varices of 50 percent at 4 years in the control arm, to be reduced to 30 percent in the timolol arm. So far our observed rates for development of varices are consistent with our planned estimates. However, we have now estimated that a minimum follow-up of 4 years (after last patient is recruited) is necessary to ensure high statistical power (80 percent at the 2-sided 0.05 level). The trial is highly significant for the promise it holds for the treatment of cirrhosis of all etiologies and for an understanding of the natural history of the disease. The four centers involved are widely renown for their studies in this area and have collaborated productively in the past, including the only published double-blind trial of propranolol in the prevention of first variceal hemorrhage in patients with cirrhosis and varices.